PROJECT ABSTRACT This project aims to develop a urine-based screening tool for prostate cancer that could affect at least 13 million men in the US who receive prostate cancer screening every year. Today, serum prostate specific antigen (PSA) remains the most commonly used screening test for prostate cancer (PCa), but the lack of specificity of PSA has led to unnecessary prostate biopsies. In addition, PCa is a heterogeneous disease ranging from indolent to life threatening or clinically significant. About 20% to 50% of men who get a positive biopsy have PCa that never grows, spreads, or harms them. Thus, there is a great need to develop better alternatives that can reliably diagnose PCa and also identify men with clinically significant prostate cancer who are most likely to benefit from early diagnosis while avoiding the over-diagnosis and overtreatment of indolent cancer. Research has found that trained dogs can distinguish patients with and without PCa by sniffing their urine. In turn, we can use these odor-producing volatile organic compounds (VOCs) as biomarkers for PCa diagnosis and risk assessment. Our preliminary data have shown that VOCs in urine are significantly different (p<0.05) between prostate cancer patients and healthy subjects. We hypothesize that pathological processes of PCa can alter the production of specific VOCs that are different and distinguishable from the VOC profile of healthy individuals. We also hypothesize that certain PCa specific VOCs are conserved across different ethnic groups, and that these VOCs are highly significant for diagnosing PCa. In this proposed study, we aim to (1) develop a urinary VOC-based screening model for PCa diagnosis; (2) create a prostate cancer risk model based on urinary VOCs for differentiating indolent from clinically significant PCa, and (3) evaluate longitudinal patterns of change of urinary VOC profiles in men with and without prostate disease. The expected outcomes and impacts of the project is to develop a non-invasive urinary VOC based model that can detect PCa with over 90 % accuracy, thus providing a better alternative to the current PSA test for PCa diagnosis. Furthermore, the VOC risk model could become a revolutionary tool with high specificity for clinically significant PCa; it has the benefit of early diagnosis while preventing over-diagnosis (i.e. finding indolent PCa that will never develop into advanced deleterious PCa) and overtreatment of indolent PCa (i.e. getting unnecessary and aggressive treatment). It would directly benefit over 13 million men receiving PSA testing in the US annually, and prevent at least 2 million unnecessary biopsies and their inherent risks (pain, bleeding, infection, death), and will reduce cost (time away from work, cost of procedure and ancillary studies) and anxiety.